[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Fwd: alioth back online, some ssh keys changed



Hi Stephen, I'm just wondering about one small thing


Look e.g. at one git repository
http://anonscm.debian.org/gitweb/?p=pkg-boinc/boinc.git

the URL field 
Just my 0.02 $ 

Gianfranco
just shows only
URL    https://alioth.debian.org/anonscm/git/pkg-boinc/boinc.git


the "git+ssh://git.debian.org/git/pkg-boinc/boinc.git" is missing.

So basically new developers won't be able to push anything to alioth if they clone a new repository from scratch.

$git push

error: Cannot access URL https://alioth.debian.org/anonscm/git/pkg-boinc/boinc.git/, return code 22
fatal: git-http-push failed


Would be nice to have a fix for this!

many thanks for your work!

have a nice weekend

Gianfranco


> Il Venerdì 22 Novembre 2013 16:57, Holger Levsen <holger@layer-acht.org> ha scritto:
> > Hi,
> 
> FTWwonder why alioth seems to be hacked now...
> 
> 
> cheers,
>     Holger
> 
> this mail forwarding is inspired by two different mails wondering why the 
> "alioth" ssh host keys have changed and after failing to find the 
> infrastructure list linked on the "developers", "misc" and 
> "other" lists on 
> http://lists.debian.org/...
> 
> thanks & kudos to all alioth admins for keeping an awesome service up! you 
> rock!
> 
> ----------  Forwarded Message  ----------
> 
> Betreff: alioth back online
> Datum: Donnerstag, 21. November 2013
> Von: Stephen Gran <sgran@debian.org>
> An: Debian Infrastructure Announce <debian-infrastructure-
> announce@lists.debian.org>
> 
> tl;dr: the Alioth service is back online, on a new host. Make sure you still
> have your data.
> 
> What went wrong?
> 
> Most of the data on vasks was stored on a RAID5. One of the disks in the RAID
> experienced a small hiccup, leading to it being thrown from the RAID, at which
> point mdadm started recovery using a hotspare. During this, another disk in 
> the
> RAID threw hard errors and we were stuck with a non-working RAID. 
> 
> Investigations showed this did not affect the /srv partition, but it affected
> the PostgreSQL partition. Due to various misconfigurations, we did not have 
> any
> recent backups of Postgres.
> 
> How it got fixed
> 
> Level-28 black voodoo Perl magic by mjt, to extract the contents of a failed
> RAID array that mdadm would barf on, allowed us to restore the database.
> 
> Alioth went from a dual-head setup, where part of the services were on
> vasks.debian.org and part were on wagner.d.o, to a single, more powerful,
> server called moszumanska.d.o. Much of the data was simply copied over, and
> some manual reconfiguration was needed to adapt to that setup.
> 
> You don't want to type that hostname, so please use the functional aliases
> (git.d.o, alioth.d.o, and so on).
> 
> What's left to do for you, dear user
> 
> · Update your SSH known keys.  You can find all debian.org ssh hostkeys
>    at [1] or on any debian.org system in /etc/ssh/ssh_known_hosts.  Or
>    you could just use the fingerprints in DNS (secured by DNSSEC).
> 
> · Check your homedir. It should have been copied over from vasks,
>    but make sure nothing is missing. Also, please take the opportunity to
>    clean up obsolete stuff, as we're often bothered by space constraints.
> 
> · Look in /mnt/home/users, which has a copy of what was on wagner,
>    and grab what you need from there. Note that this mountpoint will
>    disappear soon (in a few weeks at most), so don't delay. Please only
>    copy what's still relevant, for the disk space considerations mentioned
>    above.
> 
> · Check your scripts and websites. We remind you that Alioth is not
>    meant to be a generic hosting facility, so if you do advanced stuff
>    you're on your own.
> 
> · Use this opportunity to evaluate your needs and your usage.  Alioth
>    is not meant to be a generic hosting facility.  Probably much of what
>    is currently hosted on alioth is better suited to living somewhere else,
>    like people.d.o (or a debian.net server, or a debian.org server).  If
>    this applies to your project, please get in touch with DSA [2].
> 
> · Copy over your crontabs from vasks/wagner.
> 
> · Check your repositories.
> 
> Cheers,
> -- 
> -----------------------------------------------------------------
> |   ,''`.                                            Stephen Gran |
> |  : :' :                                        sgran@debian.org |
> |  `. `'                        Debian user, admin, and developer |
> |    `-                                    http://www.debian.org |
> -----------------------------------------------------------------
> [1] https://db.debian.org/debian_known_hosts
> [2] http://dsa.debian.org/
> 
> -------------------------------------------------------
>


Reply to: